# US 2026/0099900 A1 · Image inversion and editing using rectified flow neural networks

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0099900 A1
- Document ID: US20260099900A1
- Status: Published application
- Kind: A1
- Application number: 19/349,016
- Filed: 2025-10-03
- Published: 2026-04-09
- Earliest claimed: 2024-10-03
- Assignees: Google Inc.
- Applicants: Google LLC
- Inventors: Litu Rout; Yujia Chen; Nataniel Ruiz Gutierrez; Wen-Sheng Chu
- Web page: https://patentgrep.com/patent/US20260099900A1

## CPC

- G06T 11/60: Creating or editing images; Combining images with text
- G06T 5/60: using machine learning, e.g. neural networks
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

Systems and methods for performing image modification. In particular, the system can, using a rectified flow neural network, perform an image inversion and image editing process to generate a modified image that has been modified according to a conditioning input received by the system.

## Claims

### Claim 1

A method performed by one or more computers, the method comprising:
obtaining an original image and a conditioning input that specifies a modification to be applied to the original image;
performing an image inversion process on a representation of the original image using a rectified flow neural network to generate structured noise; and
performing an editing process using the rectified flow neural network and conditioned on the conditioning input to map the structured noise to a representation of a modified image.

### Claim 2 (depends on 1)

The method of claim 1, wherein performing an image inversion process on a representation of the original image using a rectified flow neural network to generate structured noise comprises:
initializing a state of the image inversion process to be the representation of the original image; and
updating the state of the inversion process at each of a plurality of forward iterations, each forward iteration having a corresponding forward time step, and the updating comprising, at each forward iteration:
processing an input comprising the state of the inversion process and the corresponding forward time step for the forward iteration using the rectified flow neural network to generate an unconditional vector field for the forward iteration;
generating a conditional vector field for the forward iteration;
combining the conditional and unconditional vectors fields for the forward iteration to generate a controlled vector field for the forward iteration; and
updating the state of the inversion process using the controlled vector field for the forward iteration.

### Claim 3 (depends on 2)

The method of claim 2, wherein the input comprising the state of the inversion process and the corresponding forward time step for the forward iteration further comprises a null representation that indicates that the unconditional vector field is not conditioned on a conditioning input.

### Claim 4 (depends on 2)

The method of claim 2, wherein combining the conditional and unconditional vector fields for the forward iteration to generate a controlled vector field for the forward iteration comprises combining the conditional and unconditional vector fields for the forward iteration in accordance with a controller guidance weight to generate the controlled vector field for the forward iteration.

### Claim 5 (depends on 2)

The method of claim 2, wherein updating the state of the inversion process using the controlled vector field for the forward iteration comprises:
updating the state of the inversion process using the controlled vector field for the forward iteration and corresponding noise levels for the forward iteration and a subsequent forward iteration.

### Claim 6 (depends on 5)

The method of claim 5, wherein updating the state of the inversion process using the controlled vector field for the forward iteration and corresponding noise levels for the forward iteration and a subsequent forward iteration comprises:
determining a difference between the corresponding noise level for the subsequent forward iteration and the corresponding noise level for the forward iteration;
determining a product of the controlled vector field and the difference; and
adding the product to the state of the inversion process.

### Claim 7 (depends on 2)

The method of claim 2, wherein generating a conditional vector field for the forward iteration comprises generating the conditional vector field based on a typical noise sample and the state of the inversion process.

### Claim 8 (depends on 7)

The method of claim 7, wherein generating the conditional vector field based on a typical noise sample and the state of the inversion process comprises:
determining a difference between the typical noise sample and the state of the inversion process; and
dividing the difference by a divisor that is based on the corresponding forward time step.

### Claim 9 (depends on 1)

The method of claim 1, wherein performing an editing process using the rectified flow neural network and conditioned on the conditioning input to map the structured noise to a representation of a modified image comprises:
initializing a state of the editing process to be the structured noise; and
updating the state of the editing process at each of a plurality of reverse iterations, each reverse iteration having a corresponding reverse time step, and the updating comprising, at each reverse iteration:
processing an input comprising the state of the editing process, a time step derived from the corresponding reverse time step for the reverse iteration, and a representation of the conditioning input using the rectified flow neural network to generate an unconditional vector field for the reverse iteration;
generating a conditional vector field for the reverse iteration;
combining the conditional and unconditional vector fields for the reverse iteration to generate a controlled vector field for the reverse iteration; and
updating the state of the editing process using the controlled vector field for the reverse iteration.

### Claim 10 (depends on 9)

The method of claim 9, wherein the time step derived from the corresponding reverse time step for the reverse iteration is equal to one minus the corresponding reverse time step for the reverse iteration.

### Claim 11 (depends on 9)

The method of claim 9, wherein the unconditional vector field for the reverse iteration is a negative of an output of the rectified flow neural network generated by processing the input comprising the state of the editing process, the time step derived from the corresponding reverse time step for the reverse iteration, and the representation of the conditioning input.

### Claim 12 (depends on 9)

The method of claim 9, wherein combining the conditional and unconditional vector fields for the reverse iteration to generate a controlled vector field for the reverse iteration comprises combining the conditional and unconditional vector fields for the reverse iteration in accordance with a controller guidance weight to generate the controlled vector field for the reverse iteration.

### Claim 13 (depends on 9)

The method of claim 9, wherein updating the state of the editing process using the controlled vector field for the reverse iteration comprises:
updating the state of the editing process using the controlled vector field for the reverse iteration and corresponding noise levels for the reverse iteration and a subsequent reverse iteration.

### Claim 14 (depends on 13)

The method of claim 13, wherein updating the state of the editing process using the controlled vector field for the reverse iteration and corresponding noise levels for the reverse iteration and a subsequent reverse iteration comprises:
determining a difference between the corresponding noise level for the subsequent reverse iteration and the corresponding noise level for the reverse iteration;
determining a product of the controlled vector field and the difference; and
adding the product to the state of the editing process.

### Claim 15 (depends on 9)

The method of claim 9, wherein generating a conditional vector field for the reverse iteration comprises generating the conditional vector field based on the representation of the original image and the state of the editing process.

### Claim 16 (depends on 15)

The method of claim 15, wherein generating the conditional vector field based on the representation of the original image and the state of the editing process, comprises:
determining a difference between the representation of the original image and the state of the editing process; and
dividing the difference by a divisor that is based on the corresponding reverse time step.

### Claim 17 (depends on 9)

The method of claim 9, wherein performing an editing process using the rectified flow neural network and conditioned on the conditioning input to map the structured noise to a representation of a modified image further comprises:
updating the state of the editing process at each of one or more additional reverse iterations that are after the plurality of reverse iterations, each additional reverse iteration having a corresponding additional reverse time step, and the updating comprising, at each additional reverse iteration:
processing an input comprising the state of the editing process, a time step derived from the corresponding additional reverse time step for the additional reverse iteration, and the representation of the conditioning input using the rectified flow neural network to generate an unconditional vector field for the additional reverse iteration; and
updating the state of the editing process using the unconditional vector field for the additional reverse iteration.

### Claim 18 (depends on 17)

The method of claim 17, wherein updating the state of the editing process using the unconditional vector field for the additional reverse iteration comprises:
updating the state of the editing process using the unconditional vector field for the additional reverse iteration without generating a conditional vector field for the additional reverse iteration.

### Claim 19

A system comprising:
one or more computers; and
one or more storage devices storing instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising:
obtaining an original image and a conditioning input that specifies a modification to be applied to the original image;
performing an image inversion process on a representation of the original image using a rectified flow neural network to generate structured noise; and
performing an editing process using the rectified flow neural network and conditioned on the conditioning input to map the structured noise to a representation of a modified image.

### Claim 20

One or more computer-readable storage media storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:
obtaining an original image and a conditioning input that specifies a modification to be applied to the original image;
performing an image inversion process on a representation of the original image using a rectified flow neural network to generate structured noise; and
performing an editing process using the rectified flow neural network and conditioned on the conditioning input to map the structured noise to a representation of a modified image.
